From the source
From the source
From the source
From the source
A few highlights from this month’s results → Muse Wonder ranks highest on overall quality and leads across four task types: analysis and answering, extraction, drafting, and editing. → GPT-6.2 Astra leads on comparison and review tasks, with overall quality 4% above November’s model average and cost per case 20% below average.
→ GPT-6.2 Sol delivers above-average quality at 78% below the average cost per case across models benchmarked in November. Explore the November results: https://t.co/pPnUpCuxJT
x.com